=== thekorn_ is now known as thekorn [08:02] Good morning [08:02] crimsun: oh, sorry; well, let's just drop or lower the delay then [08:03] Good afternoon pitti :) [08:04] didrocks, desrt: valac is in main; shouldn't it be? [08:04] hey RAOF, had a good weekend? [08:04] Pretty good, yaeh. [08:04] pitti, it is, indicator build-depends on it [08:04] to build the gir binary [08:04] am I just not fully awake yet, or do fonts look very ugly since today? [08:05] I got me a shiny new camera for my birthday :) [08:05] pitti, I didn't upgrade yet but I upgraded on saturday [08:05] looks fine there [08:07] pitti, the retracers emailing every hour is that to say they are running? ;-) [08:07] pitti, didrocks, desrt: also you should really get vala 0.8.0, it fixes *many* bugs and is the start of a stable release series (while 0.7) wasn't [08:07] * seb128 cleans some 118 retracer emails [08:07] seb128: sorry about that; no clue why this now happens, but I'll look into this today [08:07] hey slomo [08:08] hi seb128 :) [08:08] slomo, it's month after new source freeze I think buzztard can wait next cycle [08:08] ok, that's fine :) [08:08] slomo, we have higher importance things to work on one week before lucid freeze than getting new untested sources and making sure they do something useful ;-) [08:09] seb128: I was barely at home during WE, so I couldn't fix it in the last two days [08:09] pitti, yeah that's ok, I'm just teasing you, easy enough to filter out [08:09] pitti, if you really want to spam you are not trying enough, you should change some letters in the title at least ;-) [08:10] re*t*racer rEtraccer etc [08:10] ;-) [08:10] d3AR Sebastien, R E Tr4cers are 80% off noW !!!11!! [08:14] good morning [08:16] bonjour didrocks [08:16] Guten Morgen pitti, how was your week-end? [08:16] we spent all Saturday driving around, to Munich and back [08:17] lut didrocks [08:17] pitti: have you been able to move everything you wanted to Munich? [08:17] my wife will be doing an internship there over the summer, so we moved over some of her stuff and already went to the work place to see where it is [08:17] salut seb128 ;) [08:18] didrocks: yes; wasn't all that much, but we filled up the car quite well [08:18] didrocks: (Renault Kangoo delivery car) [08:19] oh :) yeah, remember about here moving to Munich for an internship, and that you will live in the two places in the meanwhile. [08:20] Living in two places at once sounds tiring. :) [08:21] hey RAOF [08:21] Hey ho, didrocks [08:22] RAOF: yeah, I'll have to switch places several times during summer [08:31] seb128, morning! Would it be possible to fix https://bugs.edge.launchpad.net/ubuntu/+source/hellanzb/+bug/549234 ? Current package is unusable and a sync should fix it [08:31] Launchpad bug 549234 in hellanzb "Please sync hellanzb (0.13-6) from debian unstable" [High,Confirmed] [08:33] cassidy, hey, adding to my list of things to look at today [08:34] cool; thanks! [09:20] seb128: ok, warning fixed [09:22] pitti, danke [09:22] seb128: it's a silly warning from python-pkgresources, FYI [09:22] seb128: I dropped our local launchpadlib from PYTHONPATH in "environ" [09:22] so that it's using the system packages [09:22] should do for now [09:24] hello [09:36] didrocks, done! New debdiff attached :) [09:36] nigelb: thanks, adding to my TODO list :) [09:37] didrocks, thank you :) [09:43] robert_ancell, hey [09:44] seb128, hery [09:44] hey [09:44] ara: who should I poke re my comment on bug 109156 [09:44] Launchpad bug 109156 in metacity "Resizing a nautlius window casues _NET_WM_MOVERESIZE window manager warnings" [Low,Triaged] https://launchpad.net/bugs/109156 [09:44] robert_ancell, how are you? [09:45] seb128, good, just arrived in Taipei this morning [09:45] * ara opens the bug report [09:45] robert_ancell, oh? what are you doing there? [09:45] * czajkowski has had a weekend of a broken lucid and buttons hiding till jcastro told me how to get them to reappear [09:45] seb128, working out of the office here for a week [09:46] czajkowski, your comment should be another bug report [09:46] ara: wasn't sure as the error is the same [09:46] czajkowski, specifically about the error message [09:47] ok [09:47] czajkowski, you are complaining about the error message not being appropriate, aren't you? [09:47] yes [09:47] and also still getting the error. [09:47] robert_ancell, oh ok, how is Taipei and the office ? [09:48] seb128, a very nice view :) much more fun flying north-south too so not jetlagged [09:48] hello everyone [09:48] czajkowski, sure, that's why you should be opening a new one only about the error message [09:48] robert_ancell, hehe [09:49] robert_ancell, is the view better than in London? [09:49] ara: ok [09:49] the view is less interesting than london but Taipei 101 is a cooler building. And a _lot_ further up :) [09:49] ara: what package should I add it to ? [09:49] Level 46 [09:50] czajkowski, metacity, I think [09:50] seb128, ^, you agree? [09:50] ara, about? [09:50] what error? [09:50] seb128, see czajkowski's comment on bug 109156 [09:50] Launchpad bug 109156 in metacity "Resizing a nautlius window casues _NET_WM_MOVERESIZE window manager warnings" [Low,Triaged] https://launchpad.net/bugs/109156 [09:50] I'm not sure [09:51] would require debugging [09:51] just pick one ;-) [09:51] that seems a low importance issue to me, I've other things I want to look at before lucid so I will let this one for triagers ;-) [09:51] seb128, :D [09:51] ara: cheers for the help 561294 [09:52] seb128: I'm sure it's low, just for a new end user to see the message that something sucks really isn't at all helpful :) [09:53] never saw that error on any install I did [09:54] yeah I find annoying errors :( [09:54] usually just me who has them too [09:59] hey seb128 - do you want to sponsor g-c-c for me? :) [09:59] nigelb: you didn't debdiff against the last version. Not a big deal there (only adapting debian/changelog), but always ensure you have the latest version :) [10:00] didrocks, ouch. I didn't know there was another upload [10:00] nigelb: apt-cache show/policy is your friend :) [10:00] nigelb: no pb, testing it now [10:00] didrocks, oh thanks :) [10:02] chrisccoulson: hey [10:02] hey seb128, how are you? [10:02] chrisccoulson: sure, thanks a lot for fixing this issue, what is taking locking there just for learning? [10:02] chrisccoulson: I'm good thank you, you? [10:03] hey chrisccoulson, you rock ;) [10:03] hey chrisccoulson, good morning [10:03] seb128 - the issue is that gtk_dialog_run takes the lock when exiting from the recursive main loop, and then it gets taken again when dispatching another event from the main loop [10:04] chrisccoulson: I don't undestand how g_timeout calling a function take the lock and not gdk_threads_add_timeout [10:04] pitti, will you be SRU-ing bug 532852 ? (you fixed it last week, I just closed it for lucid) [10:04] Launchpad bug 532852 in policykit-1 "pkexec information disclosure vulnerability" [Undecided,Confirmed] https://launchpad.net/bugs/532852 [10:04] so, using gtk_threads_add_timeout rather than g_timeout_add guards the whole callback with gdk_threads_enter/gdk_threads_leave, and makes sure it ends up in a consistent state [10:04] chrisccoulson: oh, I see [10:04] nigelb: I wasn't going to (I closed the karmic task); is there a pressing reason for this? [10:05] nigelb: oh, right, that's the other bug, sorry [10:05] nigelb: I thought about the crash on invalid pid [10:05] hehe :) [10:07] chrisccoulson: so, the gdk_threads_enter was called twice? (didn't noticed that when adding some printf). Not sure to follow you there between the different between g_add_timeout and gdk_threads_add_timeout [10:07] didrocks - events dispatched from g_timeout_add don't have the GDK global lock (ie, gdk_threads_enter hasn't been called) [10:08] whereas gdk_threads_add_timeout gives you the global lock [10:09] chrisccoulson: oh ok, so, the callback function as discaring in the first case the global lock. Understood [10:09] thanks :) [10:09] yw [10:10] didrocks - this explains some of it i think: http://library.gnome.org/devel/gtk-faq/stable/x481.html [10:10] chrisccoulson: thanks, will have a look [10:10] chrisccoulson: did you have a chance to look at the dnd crasher too will you were at it? [10:11] seb128 - i didn't get a chance. i finished quite late on friday in the end ;) [10:11] chrisccoulson: no problem, I will sponsor that already for now ;-) [10:11] thanks [10:12] thank *you* [10:13] Hi there [10:15] hey pitti, how are you btw? (sorry, i missed your comment above) ;) [10:15] chrisccoulson: I'm good, thanks! [10:16] how about you? [10:16] lut baptistemm [10:16] pitti - quite tired. i had to work a bit at the weekend, as the NSS security update in karmic caused 2 regressions [10:16] urgh [10:16] chrisccoulson: perhaps take off the afternoon to get some rest? [10:16] yes. we all suffered that way ;) [10:17] (once) [10:17] * pitti hugs asac and chrisccoulson [10:17] hehe [10:17] but they're both fixed now thanks to jdstrand and asac :) [10:17] * chrisccoulson hugs pitti [10:17] hello seb128 [10:17] * asac hugs pitti and chrisccoulson .... and seb128 ;) [10:17] * chrisccoulson hugs asac and seb128 too [10:17] * seb128 hugs asac [10:17] * seb128 hugs chris [10:17] * seb128 hugs chrisccoulson [10:18] bah, namespace conflicts on this channel :p [10:18] pitti, "usb 5-2: usbfs: USBDEVFS_CONTROL failed cmd rhythmbox rqt 128 rq 6 len 1024 ret -84" [10:18] lol [10:18] do you know what those USBDEVFS_CONTROL failed are? [10:18] -ETOOMANYCHRIS [10:18] seb128: I've never seen those, I'm afraid; is that libmtp? [10:18] bug #559892 [10:18] Launchpad bug 559892 in rhythmbox "rhythmbox tries to control apple's mighty mouse" [Low,New] https://launchpad.net/bugs/559892 [10:19] seb128: usbdevfs is /dev/bus/usb/ [10:19] lol [10:19] pitti, another of those "rhythmbox hijack what is should not" [10:19] is -> it [10:20] combined with bug #557623 the new mtp code creates trouble [10:20] Launchpad bug 557623 in libusb "Valgrind invalid read error in usb_parse_descriptor()" [Undecided,New] https://launchpad.net/bugs/557623 [10:20] I'm pondering turning mtp off by default in rhythmbox for lucid [10:20] seb128: udevadm info --exportdb output might be useful [10:20] pitti, I've asked that on the bug, thanks [10:20] not to mention that we use an ages old libusb [10:21] seb128: hmm, that'd be a pity.. it'd be worth checking if there's a /dev/libmtp* link for it [10:21] so nobody is going to fix it for us [10:21] seb128: i. e. if it's libmtp recognizing too many devices, or Rhythmbox tryign to grab non-MTP devices (which I suspect) [10:21] pitti, what does create the entry? [10:21] the latter is the same problem with gphoto cams [10:21] seb128: /lib/udev/rules.d/45-libmtp8.rules [10:21] well [10:21] SYMLINK+="libmtp-%k" [10:22] one issue is that libmtp crashes in libusb [10:22] cf the bug I just pointed [10:22] I'm not sure how to deal with that but it crashes rhythmbox for lot of users [10:22] since rhythmbox does a libmtp_detect at init [10:23] the second issue is that it hijack some non mtp devices which I don't understand [10:23] the cameras case is special since those a ptp devices which is sort or mtp-ish too [10:23] still a bug [10:23] right, but they don't generate a libmtp dev symlink [10:23] at least not for mine [10:23] but I don't get how it hijack usb key mounting or mouse for some users [10:23] and yet RB tries to grab it [10:24] right, mtp-detect returns things for those though [10:24] I have a feeling that RB iterates over all RB devices and does some not-tight-enough checks [10:24] which rhythmbox doesn't handle correctly [10:24] right [10:24] I will look to that today [10:25] still it will not fix the crasher [10:25] the crasher is with real MTP devices? [10:26] no [10:27] I get the invalid read on my d630 with no device attached to the computer [10:27] just calling the libmtp detect function is tnough [10:29] seb128: so RB crashes right away on your system? [10:29] pitti, no, it's an invalid read, it's like lottery [10:30] depends what you hit with your pointer, it might just not crash [10:31] I do get the invalid read under valgrind though and crash report we get indicates it's one of the most common lucid crashes for rhythmbox [10:31] * pitti tries the reproducer [10:33] seb128: hm, I ran that with "valgrind testmtp", and no errors [10:33] seb128: do I need any option? [10:33] no [10:34] $ valgrind ./device [10:34] ==8536== Memcheck, a memory error detector [10:34] ... [10:34] ==8536== Invalid read of size 1 [10:34] ==8536== at 0x4085E08: usb_parse_descriptor (descriptors.c:42) [10:34] ==8536== by 0x40860BF: usb_parse_configuration (descriptors.c:238) [10:34] hm, then I can't reproduce this, I'm afraid [10:34] it might well be hardware specific [10:34] looking to the libusb code it seems case where it fails to parse to usb infos [10:36] seb128: would it help at all to build RB and/or libmtp against libusb-1.0? [10:36] I can try if libusb1 has this bug [10:36] but it seems late for a such change [10:36] they seem to have a different API, though [10:36] teuf said libusb1 has a compat layer [10:36] seb128: right, but if -1.0 works, we might backport a fix from upstream [10:37] we maybe don't build the compat lib though [10:41] seb128, if you have time, could you please add an endorsement to my application at https://wiki.ubuntu.com/AraPulido/PerPackageUploadApplication [10:41] seb128, comments are also welcome [10:41] ara, ok, will do! [10:41] seb128, thanks! [10:41] you're welcome [10:43] ah, persia has suggested that I could have endorsement for bluetooth related packages [10:44] he told that he will endorse but I wonder if I needed someone else? [10:44] perhaps someone here could do that? [10:46] I'm not sure, I didn't sponsor enough of your updates to say [10:46] at least you fixed some garbage of mine :) [10:47] right, which make me think you are sometime not careful enough ;-) [10:47] let's maybe wait next cycle [11:13] seb128: thanks for subscribing me to all the volume handling bugs (in nautilus, etc.) [11:13] * pitti can leech from seb128's awesome bug triaging and concentrate on the ones that he can help with [11:13] pitti, is that really welcome or do you want me to stop doing that? ;-) [11:13] seb128: no, I'm serious [11:13] good [11:13] * seb128 hugs pitti [11:13] I don't find enough time to read all the gvfs etc. bugs myself [11:14] * pitti hugs seb128 [11:22] ok, I finished email backlog before lunch today [11:29] pitti, I didn't see your new comment on bug #559723 and reassign to udisks, do you want me to reassign it back? [11:29] Launchpad bug 559723 in udisks ""phantom" audio cd" [Low,Confirmed] https://launchpad.net/bugs/559723 [11:30] pitti, the udisks log seems boggus some thousand tracks and 96 audio ones too [11:30] seb128: it's probably udev, let me look [11:31] seb128: ah; I'll wait for his udev log and cdrom_id output, and reassign to udev if appropriate [11:31] udisks is fine for now [11:31] ok good [11:31] * pitti -> doctor and lunch, bbl [11:31] hum, lunch! [11:33] chrisccoulson: do you have some progress on bug 536766? [11:33] Launchpad bug 536766 in gnome-user-share "Personal file sharing preferences dialog does not offer to install needed packages" [Low,Triaged] https://launchpad.net/bugs/536766 [11:33] pitti: see you [11:33] seb128: enjoy === MacSlow is now known as MacSlow|lunch [12:28] didrocks: hi , re: the cheese apport hook. the log file now attaches as a .log or still a .log.gz ? [last i heard nigel was having trouble getting that done..] [12:30] vish: hey, from the source it's a .log now, but I'll have to check again once the crash will be fixed :) [12:31] didrocks: ah, then probably not , lp does the compression :( iirc , he mentioned he would work on it again to fix that part for Maverick [12:32] vish: ok, maybe we can keep it that way for lucid, even if it's not nice for us to get that and see for maverick with LP guys (having an apport hook is still better than nothing) [12:32] yup :) [12:49] hello. maybe could anyone take a look, please, at #524938 and at some rationale, why this package should be fixed, not removed - https://bugs.launchpad.net/ubuntu/+source/libopensync-plugin-syncml/+bug/524938/comments/10 [12:49] Launchpad bug 524938 in libopensync-plugin-syncml "Remove binary "opensync-plugin-syncml" from lucid" [Undecided,New] [13:23] seb128: hi hi [13:24] hey gicmo! [13:24] how are you? [13:25] pretty good [13:25] in Munich at the University [13:25] you will be pleased to know that everybody here uses ubuntu [13:25] ;-) [13:27] gicmo, waouh! [13:27] gicmo, when is the gsetting hackfest? [13:29] seb128, this week, coulnd't make it, but will be helping out remotely on Wed, Thur, Fr. [13:29] gicmo, oh ok === ogra_ is now known as ogra [13:33] seb128: I hope lucid comes out on time, here everything is on hardy and is eager to upgrade to next LTS [13:33] gicmo, Ubuntu is like GNOME, it's always out on time ;-) [13:34] now why did i read "of" instead of "on" above === MacSlow|lunch is now known as MacSlow [13:35] seb128: hehe ;-) [13:35] ogra: heh freudian reading error? [13:35] (-: [14:06] gicmo, Ubuntu is like GNOME, it's always out on time ;-) [14:06] ups [14:06] gicmo, sorry, focus issue [14:31] rickspencer3, hey [14:33] hi seb128 [14:33] pitti, seb128: a recommends should be enough to pull something onto the CD right? [14:33] kenvandine, hi [14:33] and good morning [14:33] kenvandine: yes [14:33] :) [14:33] kenvandine, yes [14:33] hey rickspencer3, hey kenvandine [14:34] pitti, python-indicate isn't on the CD but is a recommends [14:34] if the something is not in universe... [14:34] un oh [14:34] damn [14:34] pitti, can that get promoted? [14:35] kenvandine: technically yes; practically, why? [14:35] * pitti waves with FF [14:35] cause the "setup broadcast accounts" in the message menu never goes away [14:35] unless that gets installed [14:35] * kenvandine doesn't know how nobody (including me) noticed this yet! [14:36] kenvandine, perhaps because desktopcouch has never succeeded in synching your gwibber account data without crashing [14:36] hahaha [14:36] true that [14:36] so no one has been able to test the SFTS experience [14:36] ^not kidding actually [14:36] I wanted to talk about this today [14:36] now that desktopcouch actually works out of the box :) [14:37] kenvandine: perhaps it should be a depends then? [14:37] over the whole weekend on my netbook, desktopcouch never succeeded in copying down all of my databases [14:37] pitti, i guess your right [14:37] since the messaging indicator kind of goes screwy without it [14:37] rickspencer3, eek [14:37] got logs? [14:37] https://bugs.launchpad.net/bugs/530541 [14:37] Launchpad bug 530541 in desktopcouch "desktopcouch-service crashed with RuntimeError in run_couchdb()" [High,Triaged] [14:38] kenvandine, this seems rather widespread [14:38] ~/.cache/desktop-couch/log/desktop-couch-replication.log [14:38] yeah, that one [14:38] is there a way for me to see what couch databases are in the cloud atm? [14:38] humm [14:38] rickspencer3, sort of [14:39] anyway, SFTS will never work if the desktopcouch just crashes after it starts up [14:39] http://www.freedesktop.org/wiki/Specifications/desktopcouch/Documentation/Troubleshooting?action=AttachFile&do=view&target=ubuntuone-couchdb-query [14:39] download that script [14:39] you can query your dbs directly in the cloud with that [14:40] but... it currently doesn't list your dbs [14:40] if you know the names, you can just hit each one [14:40] also look at your replication log [14:40] bug #528728 [14:40] Launchpad bug 528728 in totem "Totem BBC plugin cannot connect" [Low,Confirmed] https://launchpad.net/bugs/528728 [14:41] hello everyone [14:41] kenvandine, ok, I'll look into that later today [14:41] in the meantime, what is the status of getting run_couchdb fixed before we, you know, release? [14:42] i think that is one of the bugs chad thinks is fixed and is looking for testers for [14:42] rickspencer3, willing to test from a ppa? [14:42] kenvandine, sure [14:42] let me get the details and confirm he thinks this is related [14:42] * kenvandine thinks he should comment to that affect on the bug [14:42] did pitti ack adding python-indicate as a depends? [14:43] it needs a MIR first [14:43] i think since he suggested it :) [14:43] pitti, the source is already main [14:43] but if it's really necessary for correct funcionality, it should be a depends, yes [14:43] it comes from libindicate [14:43] ah [14:43] * pitti promotes then [14:43] so no MIR [14:43] thx [14:43] i'll add as a depends anyway [14:43] technically gwibber works, but it breaks the experience for messaging menu [14:43] kenvandine: no, the source is indicate-python [14:44] oh? [14:44] damn [14:44] * kenvandine thought that was built from the same source [14:44] ok, i'll get an MIR together [14:45] next cycle we should merge that inline with libindicate :) [14:45] then tedg doesn't forget about it... and learns to love python [14:45] kenvandine: If you can fix indicator-application so it builds twice from the same build, I'd love python more :) [14:46] For Maverick will we be able to just use the GIR file for Python? [14:47] dunno [14:47] anything that uses GIR doesn't build for me now... :) [14:47] * tedg wants that so bad. This building multiple Python versions stuff is *so* stupid. [14:51] pitti, bug 561508 [14:51] Launchpad bug 561508 in indicate-python "[MIR] indicate-python" [Undecided,New] https://launchpad.net/bugs/561508 [14:52] Hey pitti. Is there anyway to manage bugs assigned to a package (i.e. likewise-open) without joining bug squad? Things like importance, won't fix, etc... no rush. Just curious, [14:52] tedg, in reality we can drop that for now [14:52] since we only care for one version... it was an issue in karmic for some reason [14:52] * kenvandine can't recall [14:52] tedg, but yes, if we can get GIR stuff instead... yay! [14:53] kenvandine: I think we can. The Python GIR hackfest is this week, eh? [14:54] coffeedude: for upstream tasks you should already be able to; but for Ubuntnu packages you need to be in bug squad, I think; can you please talk to bdmurray or pedro to add you? [14:54] pitti, Will do. Thanks. [15:00] kenvandine: are you working on proxy support for gwibber for lucid? [15:00] popey, no :( [15:00] :( [15:00] we have a patch adding libproxy support, but it requires a newer libproxy that is in lucid [15:00] and it is too late to update something like that [15:00] when I start gwibber behind a proxy I get apport crashes for gwibber, gwibber-service and desktop-couch immediately [15:00] popey, it will go in real soon though for lucid+1 [15:01] popey, humm... crashes? [15:01] not even an SRU? [15:01] yup [15:01] barfaroony [15:01] popey, it is a pretty big jump, libproxy 0.2.3 to 0.4.0 [15:01] should break anything, but might require some rebuilds [15:01] popey, we will upload the newer libproxy to the gwibber ppa though for folks using the newer gwibber [15:02] we might want to consider the new libproxy [15:02] seb128, oh? [15:02] it's late but not used by a lot of things [15:02] i would love to update it [15:02] telepathy guys raised the issue [15:02] the version we have is year old and buggy [15:02] libsoup-gnome2.4-1 [15:02] I wanted to maybe raise it tomorrow in the meeting or discuss it with pitti later when he's done dealing with lucid breakages [15:02] is the only thing that concerned me [15:03] right [15:03] seb128, if you can get that updated... i'll fix the proxy issue and make a ton of people happy :) [15:03] +1 :) [15:03] the proxy part is probably not used frequently in normal desktop use [15:05] seb128, yeah, and the developers said there is no real api breakages, just additions [15:05] with one exception, i think something no longer returns NULL when it used to [15:05] kenvandine, is the proxy code used transparently by any libsoup client or is that something you need to use specifically? [15:06] i need specifically [15:06] for gwibber [15:06] to do anything useful for proxy support in gwibber, we need 0.4.x [15:06] well the context was rather to evaluate how much impact it can have on libsoup users [15:07] ie if it's an opt in and gvfs etc don't use it it has pretty low chance to break anything [15:07] it = the update [15:14] pitti, is there something wrong with the apport.hookutils.attach_hardware code? I get a traceback while running a hook [15:14] didrocks, looking at your comments. It has something to do with ^ [15:18] nigelb: I need some more details about that (like the backtrace) before I can answer. It generally works fine [15:18] pitti, https://bugs.launchpad.net/ubuntu/+source/cheese/+bug/542091/comments/15 [15:18] Launchpad bug 542091 in cheese "Add apport hook for cheese" [Wishlist,Triaged] [15:19] I did see it when I tested my self, but ignored it then [15:20] nigelb: ah, you aren't supposed to use the return value (it's None) [15:20] nigelb: just call attach_hardware(repoert) [15:20]